GLP-1 Muscle Loss Prevention: The Evidence-Based Protocol

Every weight-loss intervention — diet, surgery, GLP-1 therapy — takes some lean mass along with the fat. The published number for tirzepatide in the SURMOUNT-1 DEXA substudy (Look 2025^[2]) is roughly −33.9% fat mass and −10.9% lean mass at week 72 on the 10 mg arm, with lean mass making up about a quarter of total body-weight loss. The same fraction shows up in older diet-only studies (Cava 2017^[4]) and the resistance-training meta-analyses (Sardeli 2018^[5]). The good news is that the protocol to minimize lean-mass loss has been replicated for two decades: protein at the high end of the evidence range (Krieger 2006^[6], Phillips 2016 ^[7]), resistance training three days per week (Sardeli 2018^[5], Longland 2016^[8]), and — for older or higher-risk patients — DEXA at baseline and again at the end of titration. This article walks through what the evidence actually says and the practical protocol it implies.

The honest summary

• Lean-mass loss is real but not catastrophic. In the SURMOUNT-1 body composition substudy (Look 2025 ^[2]), tirzepatide 10 mg produced −33.9% fat mass and −10.9% lean mass at week 72; the lean fraction of total weight loss was about 25%. That is the same fraction seen in dietary weight loss without medication (Cava 2017^[4]).
• Protein dose has a published threshold. Krieger 2006 (AJCN meta-regression^[6]) and Phillips 2016^[7] converge on roughly 1.6 g/kg per day as the minimum intake associated with preserved fat-free mass during energy restriction; intakes up to 2.0–2.4 g/kg add small additional benefit during intense exercise (Longland 2016^[8]).
• Resistance training is the lever. Sardeli 2018^[5] (meta-analysis of older adults in a caloric deficit) found resistance training abolished the lean-mass loss otherwise seen with diet alone. Two to three sessions per week of compound lifts is the typical protocol.
• Screen for sarcopenic obesity in older patients. The ESPEN/EASO consensus (Donini 2022^[11]) and EWGSOP2 (Cruz-Jentoft 2019^[12]) provide screening criteria; patients age ≥ 65 or with low baseline grip strength deserve a DEXA at baseline and formal sarcopenia screening before starting a GLP-1.

What the SURMOUNT-1 body composition substudy actually showed

SURMOUNT-1 (Jastreboff 2022 NEJM^[1]) randomized 2,539 adults with obesity to tirzepatide 5, 10, or 15 mg weekly or placebo for 72 weeks. Mean total body-weight loss was −15.0%, −19.5%, and −20.9% on the three active arms vs −3.1% on placebo. The pre-specified body composition substudy (Look 2025, Diabetes Obesity & Metabolism^[2]) added serial DEXA to 160 participants and reported:

• Fat mass: −33.9% from baseline at week 72 on tirzepatide 10 mg.
• Lean mass: −10.9% from baseline at week 72 on tirzepatide 10 mg.
• Lean fraction of total weight loss: about one quarter — in line with the “~25%” benchmark from diet-only weight loss.
• Fat-to-lean ratio improved overall: the ratio of fat mass to lean mass dropped substantially because the absolute fat loss was three times the absolute lean loss.

The headline interpretation: tirzepatide does not produce disproportionately more lean-mass loss than non-pharmacological weight loss of similar magnitude. It produces the same fraction of lean loss against a much larger fat loss, so the relative body composition improves.

STEP-1 and what the semaglutide picture looks like

STEP-1 (Wilding 2021 NEJM^[3]) randomized 1,961 adults to semaglutide 2.4 mg weekly or placebo and reported −14.9% total body weight at week 68. A DEXA substudy was published as part of the STEP program with a similar pattern: fat mass dropped more than lean mass, but lean mass did account for a clinically meaningful fraction of the loss. The directional message matches SURMOUNT-1: the lean fraction of weight loss with semaglutide and tirzepatide is in the same range as diet-induced weight loss, around a quarter to a third — not the 50%+ that early speculation feared.

The protein evidence: 1.6 g/kg is the floor

Three independent lines of evidence converge on the same protein dose for lean-mass preservation during a caloric deficit.

Krieger 2006^[6] meta-regressed 87 studies of energy-restricted diets and found that protein intake above ~1.05 g/kg was associated with significantly less fat-free mass loss; the dose-response continued upward. Phillips 2016^[7] reviewed the leucine-threshold and muscle protein synthesis literature and landed on roughly 1.6 g/kg per day as the practical ceiling of incremental benefit for healthy adults. Longland 2016^[8] randomized young men to either 1.2 or 2.4 g/kg during a 40% energy deficit plus high-intensity training and found the high-protein arm gained ~1.2 kg of lean mass while losing ~4.8 kg of fat — the low-protein arm preserved but did not gain. Wycherley 2012 ^[9] meta-analyzed 24 RCTs of higher-protein energy-restricted diets and confirmed reduced fat-free mass loss and greater fat loss vs standard-protein controls.

The practical translation for GLP-1 patients is 1.6–2.0 g/kg of current body weight per day. A 90 kg adult on Wegovy targets 144–180 g of protein daily. That target is hard to hit when a GLP-1 has cut appetite by a third, which is exactly why the protein side of the protocol fails more often than the training side.

The resistance-training evidence: two to three sessions per week

Sardeli 2018^[5] meta-analyzed six RCTs of resistance training during caloric restriction in obese older adults. Diet alone produced lean-mass loss averaging about 5% of baseline; diet plus resistance training produced effectively no lean-mass loss. The effect was robust across protocols using 2–3 sessions per week of moderate-intensity lifting. Cava 2017 ^[4] reviewed the broader literature and reached the same conclusion: the single most effective countermeasure to weight-loss-induced muscle loss is resistance exercise.

The training protocol that supports the published outcomes is unglamorous: compound multi-joint lifts (squat, hinge, press, row, pull), 2–3 sessions per week, progressive overload on a 3–6 rep-range working set with 1–2 backoff sets, and at least 48 hours between sessions targeting the same muscle group. The literature does not require a barbell — machine-based or dumbbell-based programs produce similar outcomes in the meta-analyses.

Creatine, vitamin D, and the supportive supplements

Creatine monohydrate 3–5 g/day is the only supplement with a position-stand-level evidence base for augmenting lean-mass preservation alongside resistance training (Kreider 2017, JISSN^[10]). The position stand reviewed several hundred studies and concluded creatine is safe, effective for resistance-training adaptation, and especially useful in older adults at sarcopenia risk — the exact population most GLP-1 patients sit inside.

Vitamin D sufficiency (serum 25-OH-D in the 30–50 ng/mL range) supports both muscle function and bone density; weight loss accelerates bone-mineral-density decline, so a baseline 25-OH-D measurement plus 1,000 mg calcium and 800–2,000 IU vitamin D daily is the usual floor. Omega-3 at 2–3 g/day has modest evidence in older adults; HMB evidence is weaker and we do not recommend it as first-line.

Magnitude: lean mass as a fraction of total weight loss

Sarcopenic obesity: who needs DEXA at baseline

The ESPEN/EASO sarcopenic obesity consensus (Donini 2022, Clinical Nutrition^[11]) and the EWGSOP2 sarcopenia consensus (Cruz-Jentoft 2019, Age and Ageing^[12]) define sarcopenia by reduced muscle strength (grip dynamometry or chair-rise time) confirmed by reduced muscle quantity (DEXA-derived appendicular lean mass index). For GLP-1 candidates, a baseline DEXA is warranted in any of:

• Age ≥ 65 — baseline sarcopenia prevalence is high.
• Grip strength below sex-specific cutoffs (< 27 kg men, < 16 kg women per EWGSOP2 ^[12]).
• Chair-rise time ≥ 15 seconds for 5 repetitions.
• Recent unintentional weight loss or rapid prior loss on another agent.
• Planned aggressive titration to maximum dose with target TBWL > 20%.

For lower-risk patients, bioelectrical impedance analysis (BIA) on a quarterly cadence is a reasonable substitute. BIA is less accurate at a single time point than DEXA, but within-patient change over time is informative. Clinically meaningful red flags are a > 15% drop in appendicular lean mass at any interim measurement, or a measured grip strength decline more than 10% from baseline.

The practical protocol

1. Protein target: 1.6–2.0 g/kg of current body weight per day. Anchor to the lower end if tolerating; push to 2.0 g/kg if you are on a high dose, older, or measurably losing lean mass. Use whey or casein before resistance sessions; spread across 3–4 doses of 30–40 g.
2. Resistance training: 2–3 sessions per week, compound lifts. Squat or leg press, hinge (deadlift or hip thrust), horizontal push, horizontal pull, vertical push, vertical pull. 3 working sets at 6–12 reps, progressive load.
3. Creatine monohydrate 3–5 g/day. Any time of day. No loading phase required.
4. Vitamin D + calcium. Baseline 25-OH-D; repeat at 6 months. 1,000 mg calcium and 800–2,000 IU vitamin D daily for most patients.
5. DEXA at baseline and at end of titration (~week 36) for any patient meeting the sarcopenic-obesity screening criteria above. Quarterly BIA for everyone else.
6. Re-evaluate at 6 months. If appendicular lean mass index has dropped > 15% or grip strength has fallen > 10%, increase protein, add a third resistance session, and consider slowing the dose ladder.

Related research and tools

• Exercise pairing on a GLP-1 — the resistance-training half of the protocol in depth
• Best protein powders for GLP-1 weight loss — whey vs casein vs plant blends, leucine content per serving
• Creatine on a GLP-1 — the Kreider position stand and what 3–5 g/day actually does
• Tirzepatide and lean mass — the SURMOUNT-1 body composition substudy in detail
• Ozempic muscle loss — semaglutide-specific protection strategies
• What to eat on a GLP-1 — how to hit 1.6 g/kg when appetite is cut by a third
• GLP-1 protein calculator — calculate your daily target by current weight
• GLP-1 exercise pairing configurator — build a 2–3 session weekly template